Shuqing Zhang

IDENTIFYING WETLAND CHANGE IN CHINA'S SANJIANG PLAIN USING REMOTE SENSING

Maximum likelihood supervised classification and post-classification change detection techniques were applied to Landsat MSS/TM images acquired in 1976, 1986, 1995, 2000, and 2005 to map land cover changes in the Small Sanjiang Plain in northeast China. A hotspots study identified land use changes in two National Nature Reserves. These were the Honghe National Nature Reserve (HNNR) and the Sanjiang National Nature Reserve (SNNR). Landscape metrics were used in both reserves to identify marsh landscape pattern dynamics. The results showed that the Small Sanjiang plain had been subject to much change. This resulted from direct and indirect impacts of human activities. Direct impacts, resulting in marsh loss, were associated with widespread reclamation for agriculture. Indirect impacts (mainly in HNNR) resulted from alterations to the marsh hydrology and this degraded the marsh ecosystem. Marsh landscape patterns changed significantly due to direct impacts in SNNR between 1976 and 1986 and again between 2000 and 2005, and, in HNNR between 1976 and 1986. Indirect impacts in HNNR after 1986 appeared to cause little change. It was concluded that effective wetland protection measures are needed, informed by the change analysis.

A variable precision rough set approach to the remote sensing land use/cover classification

Nowadays the rough set method is receiving increasing attention in remote sensing classification although one of the major drawbacks of the method is that it is too sensitive to the spectral confusion between-class and spectral variation within-class. In this paper, a novel remote sensing classification approach based on variable precision rough sets (VPRS) is proposed by relaxing subset operators through the inclusion error @b. The remote sensing classification algorithm based on VPRS includes three steps: (1) spectral and textural information (or other input data) discretization, (2) feature selection, and (3) classification rule extraction. The new method proposed here is tested with Landsat-5 TM data. The experiment shows that admitting various inclusion errors @b, can improve classification performance including feature selection and generalization ability. The inclusion of @b also prevents the overfitting to the training data. With the inclusion of @b, higher classification accuracy is obtained. When @b=0 (i.e., the original rough set based classifier), overfitting to the training data occurs, with the overall accuracy=0.6778 and unrecognizable percentage=12%. When @b=0.07, the highest classification performance is reached with overall accuracy and unrecognizable percentage up to 0.8873% and 2.6%, respectively.

Improved land cover mapping using random forests combined with landsat thematic mapper imagery and ancillary geographic data.

Large area land-cover mapping involving large volumes of data is becoming more common in remote sensing applications. Thus, there is a pressing need for increased automation in the land-cover mapping process. The main objective of this research was to compare the performance of three machine learning algorithms (MLAS) for mapping wetlands in the Sanjiang Plain combined Landsat TM imagery with ancillary geographical data. Three MLAS included random forest (RF), classification and regression tree (CART), and maximum likelihood classification (MLC). Comparisons were based on several criteria: overall accuracy, sensitivity to data set size, and noise. Our results indicated that first, the random forest and CART approach can achieve substantial improvements in accuracy over the traditional MLC method. Random forest produced the highest overall accuracy (91.3 percent) the kappa coefficient 0.8943, with marsh class accuracies ranging from 77.4 percent to 90.0 percent. Secondly, the random forest method was least sensitive to reduction in training sample size, and it was most resistant to the presence of noise compared to CART and MLC. The comparison between three MLAS revealed that the random forest approach was most resistant to training data deficiencies while improved land-cover map accuracy in marsh area.

Straight road edge detection from high-resolution remote sensing images based on the ridgelet transform with the revised parallel-beam Radon transform

Roads are important basic geographical phenomena and the automatic recognition and extraction of road features from remote sensing images has many applications. However, automated road extraction from high-resolution remote sensing imagery is problematic. In recent years, many approaches have been explored for automatic road extraction, particularly involving road edge detection. Traditional edge detection operators such as the Canny or the Sobel operator are used frequently but there are serious problems of over- or underdetection, and time-consuming and complicated post-processing work is often required. In this paper, a new revised parallel-beam Radon transform (RPRT) approach is proposed. The traditional PRT can have problems with step values, resulting in false edge detection. To overcome these problems we introduced the RPRT, using the harmonic average of the pixel value in every strip of the Radon slice. An algorithm suitable for straight edge detection of roads in high-resolution remote sensing imagery was designed based on the ridgelet transform with the RPRT. The experimental results show that our algorithm can detect straight road edges efficiently and accurately, and avoid cumbersome and complicated post-processing work.

Using CropSyst to Simulate Spring Wheat Growth in Black Soil Zone of Northeast China

ABSTRACT Available water and fertilizer have been the main limiting factors for yields of spring wheat, which occupies a large area of the black soil zone in northeast China; thus, the need to set up appropriate models for scenario analysis of cropping system models has been increasing. The capability of CropSyst, a cropping system simulation model, to simulate spring wheat growth of a widely grown spring cultivar, ‘Longmai 19’, in the black soil zone in northeast China under different water and nitrogen regimes was evaluated. Field data collected from a rotation experiment of three growing seasons (1992–1994) were used to calibrate and validate the model. The model was run for 3 years by providing initial conditions at the beginning of the rotation without reinitializing the model in later years in the rotation sequence. Crop input parameters were set based on measured data or taken from CropSyst manual. A few cultivar-specific parameters were adjusted within a reasonable range of fluctuation. The results demonstrated the robustness of CropSyst for simulating evapotranspiration, aboveground biomass, and grain yield of ‘Longmai 19’ spring wheat with the root mean square errors being 7%, 13% and 13% of the observed means for evapotranspiration (ET), grain yield and aboveground biomass, respectively. Although CropSyst was able to simulate spring production reasonably well, further evaluation and improvement of the model with a more detailed field database was desirable for agricultural systems in northeast China.

Performance Evaluation of Cluster Validity Indices (CVIs) on Multi/Hyperspectral Remote Sensing Datasets

The number of clusters (i.e., the number of classes) for unsupervised classification has been recognized as an important part of remote sensing image clustering analysis. The number of classes is usually determined by cluster validity indices (CVIs). Although many CVIs have been proposed, few studies have compared and evaluated their effectiveness on remote sensing datasets. In this paper, the performance of 16 representative and commonly-used CVIs was comprehensively tested by applying the fuzzy c-means (FCM) algorithm to cluster nine types of remote sensing datasets, including multispectral (QuickBird, Landsat TM, Landsat ETM+, FLC1, and GaoFen-1) and hyperspectral datasets (Hyperion, HYDICE, ROSIS, and AVIRIS). The preliminary experimental results showed that most CVIs, including the commonly used DBI (Davies-Bouldin index) and XBI (Xie-Beni index), were not suitable for remote sensing images (especially for hyperspectral images) due to significant between-cluster overlaps; the only effective index for both multispectral and hyperspectral data sets was the WSJ index (WSJI). Such important conclusions can serve as a guideline for future remote sensing image clustering applications.

Voronoi diagram and GIS-based 3D path planning

the integration of terrain following, terrain avoidance, threat avoidance (TF/TA2) is the key technique for aircrafts to achieve low altitude penetration flight. Depending on this technique, survival ability, accuracy and diversity of aerial assault have been greatly improved. In this Paper, the optimal trajectory programming algorithm of TF/TA2 has been improved, and threat avoidance has been studied.

Effects of normalized difference vegetation index and related wavebands’ characteristics on detecting spatial heterogeneity using variogram-based analysis

Spatial heterogeneity is widely used in diverse applications, such as recognizing ecological process, guiding ecological restoration, managing land use, etc. Many researches have focused on the inherent scale multiplicity of spatial heterogeneity by using various environmental variables. How these variables affect their corresponding spatial heterogeneities, however, have received little attention. In this paper, we examined the effects of characteristics of normalized difference vegetation index (NDVI) and its related bands variable images, namely red and near infrared (NIR), on their corresponding spatial heterogeneity detection based on variogram models. In a coastal wetland region, two groups of study sites with distinct fractal vegetation cover were tested and analyzed. The results show that: 1) in high fractal vegetation cover (H-FVC) area, NDVI and NIR variables display a similar ability in detecting the spatial heterogeneity caused by vegetation growing status structure; 2) in low fractal vegetation cover (L-FVC) area, the NIR and red variables outperform NDVI in the survey of soil spatial heterogeneity; and 3) generally, NIR variable is ubiquitously applicable for vegetation spatial heterogeneity investigation in different fractal vegetation covers. Moreover, as variable selection for remote sensing applications should fully take the characteristics of variables and the study object into account, the proposed variogram analysis method can make the variable selection objectively and scientifically, especially in studies related to spatial heterogeneity using remotely sensed data.

Deduction and application of generalized Euler formula in topological relation of geographic information system (GIS)

By combining geographic information system (GIS), a new conception of “group” in which covers the complex connected region is introduced in the paper. Based on the conception of group, a generalized Euler formula and its properties are deduced and proved. The paper also describes the mathematical principles of generating topological information of polygon in GIS maps and the methods for checking up the veracity of topological relations of the map with Euler formula and general Euler formula. We have also obtained the quantitative relations among real nodes, chains, islands and groups with the formulas. At the same time, the paper introduces Whole Sphere Stereographic (WSS) projection into the GIS, and defines a new conception of “sea”. The deduction of general Euler formula and the introduction of WSS projection to GIS have developed new ways of delineating GIS topological models from plane to sphere and even constructing three dimensional (3-D) topological models.

A novel unsupervised bee colony optimization UBCO method for remote-sensing image classification: a case study in a heterogeneous marsh area

ABSTRACT Unsupervised image classification is an important means to obtain land-use/cover information in the field of remote sensing, since it does not require initial knowledge (training samples) for classification. Traditional methods such as k-means and Iterative self-organizing data analysis technique (ISODATA) have limitations in solving this NP-hard unsupervised classification problem, mainly due to their strict assumptions about the data distribution. The bee colony optimization (BCO) is a new type of swarm intelligence, based upon which a simple and novel unsupervised bee colony optimization (UBCO) method is proposed for remote-sensing image classification. UBCO possesses powerful exploitation and exploration capacities that are carried out by employed bees, onlookers, and scouts. This allows the promising regions to be globally searched quickly and thoroughly, without becoming trapped on local optima. In addition, it has no restrictions on data distribution, and thus is especially suitable for handling complex remote-sensing data. We tested the method on the Zhalong National Nature Reserve (ZNNR) – a typical inland wetland ecosystem in China, whose landscape is heterogeneous. The preliminary results showed that UBCO (overall accuracy = 80.81%) achieved statistically significant better classification result (McNemar test) in comparison with traditional k-means (63.11%) and other intelligent clustering methods built on genetic algorithm (unsupervised genetic algorithm (UGA), 71.49%), differential evolution (unsupervised differential evolution (UDE), 77.57%), and particle swarm optimization (unsupervised particle swarm optimization (UPSO), 69.86%). The robustness and superiority of UBCO were also demonstrated from the two other study sites next to the ZNNR with distinct landscapes (urban and natural landscapes). Enabling one to consistently find the optimal or nearly optimal global solution in image clustering, the UBCO is thus suggested as a robust method for unsupervised remote-sensing image classification, especially in the case of heterogeneous areas.